{#if visible}
{ if (e.key === 'Escape') handleCancel(); }}>

Settings

{#if activeTab === 'transcription'}
{:else if activeTab === 'speakers'}

Setup (one-time)

Speaker detection uses pyannote.audio models hosted on HuggingFace. You must accept the license for each model:

  1. Create a free account at openUrl('https://huggingface.co/join')}>huggingface.co
  2. Accept the license on all three of these pages:
  3. Create a token at openUrl('https://huggingface.co/settings/tokens')}>huggingface.co/settings/tokens (read access)
  4. Paste the token above and click Test & Download
{#if modelStatus === 'success'}

Model downloaded successfully. Speaker detection is ready.

{/if} {#if modelStatus === 'error'}

{modelError}

{/if}

Hint the expected number of speakers to speed up diarization clustering.

{:else if activeTab === 'ai'}
{#if localSettings.ai_provider === 'openai'}
{:else if localSettings.ai_provider === 'anthropic'}
{:else if localSettings.ai_provider === 'litellm'}
{/if} {:else}

Place GGUF model files in ~/.voicetonotes/models/ for auto-detection. The local AI server uses the OpenAI-compatible API from llama.cpp.

{:else if activeTab === 'developer'}

Opens the browser inspector for debugging. Changes take effect immediately.

{/if}
{/if}